PX

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 1of7

DECLARATION OF SAMUEL JOHNSTON
PURSUANT TO 28 U.S.C. § 1746

I, Samuel Johnston, hereby state that I have personal knowledge of the facts set forth
below and am competent to testify about them. If called as a witness, I could and would testify

as follows:
1. I am over 18 years old. I live in Kansas City, Kansas.
2 I am familiar with the process of bitcoin mining, and was looking into it on the

internet as early as 2010 to 2011. As a result of my interest, I got in touch with Sonny Vleiseides
using Skype around March 2012. I then purchased bitcoin mining equipment from BF Labs, Inc.
(“Butterfly Labs”). I ordered its FPGA mini-rig and later upgraded to its 65 nm ASIC mini-rig.

a I became employed with Butterfly Labs through my relationship with Mr.
Vleiseides in June 2013 and worked there until November 2013.

4, During my time at Butterfly Labs, I worked from its premises located at 10770 El
Monte Street. For a time, I was working in the shipping room, and in August 2013, I was
transferred to a new role as Head Burn-in Technician. In that role, I was in charge of overseeing
testing of the bitcoin mining machines that Butterfly Labs sold to consumers.

5. At the time I started working at Butterfly Labs, I was aware that the company had
made statements to consumers that its bitcoin mining machines were tested on the bitcoin testnet,
which meant that the machines would not produce any bitcoin value while being tested. From
the time I started, I observed that the machines were not in fact tested on the testnet. Instead, I
found that they were mining with the machines on the bitcoin network.

6. Butterfly Labs referred to the process of testing customer equipment by mining on
the bitcoin network as the burn in of the machines because the first ASIC machines became very

hot after being tested on the bitcoin network.

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 2 of 7

ve While I was employed, with the exception of a two-to-three-week period, all
tested machines were set up to mine bitcoins in the Eclipse Mining Consortium (“EMC”). I
believe that EMC was owned by Josh Zerlan. I was told by Mark Goodpasture at one point that
the mined bitcoins generated by the burn in process would be put into a fund to benefit
employees or charity. However, I later learned from discussions with Mark and other personnel
that there was no such charity or employee fund. Instead, the burn in process was set up to mine
bitcoins for the company’s benefit.

8. When I asked Mark why the machines were not tested on the testnet, he
responded that there was no point in doing so because the company would not make any money
from the testing. I understood, based on my conversations with Mark, that he received direction
from Sonny to test the machines in a way that generated bitcoins. I was present in at least one
brief meeting with Sonny and Mark in which it was suggested that the machines be tested on the
testnet, but Sonny and Mark agreed the company would not do so. I understood that the sole
reason for burning in the machines rather than using the testnet was to make money.

9. Bitcoins generated from burning in customer equipment went to Butterfly Labs
accounts. Initially, the machines burned into the EMC pool. At various times, the EMC pool
would go down. As a result, I set up two additional address for bitcoin wallets, including one on
my personal computer and another on blockchain.info. When the machines were tested, they
generated bitcoins though a public pool called eligius.st, and the bitcoins were sent to those two
wallets. My personal computer containing one of the bitcoin wallets was later stolen.

10. Butterfly Labs’ mining operation using equipment to be shipped to customers was
powerful. For example, in August 2013, the machines being tested were collectively hashing at a

rate of 12 terrahashes per second (TH/s). At that time, this comprised approximately 3% of the

2

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 3 of 7

hash rate of the entire bitcoin network. Based on how the bitcoin network operates, the higher
the hash rate, the higher the mining difficulty level becomes. Butterfly Labs’ mining operation
therefore made it more difficult for other miners on the bitcoin network to mine bitcoins.

11. In or around September 2013, as part of my job duties, I began setting up separate
burn in rooms for testing the bitcoin mining machines before shipping. I set up three rooms at
the El Monte premises. One of those rooms was located in a trailer in the back of the building.
Another one of the rooms was located in the shipping area to which the trailer was attached, and
another was accessible by the main hallway and attached to the packaging area.

12. The goal was to maximize the number of units that could be burned in at a time.
Mark oversaw my duties relating to burning in machines. Mark reported to Sonny about the
burn in process, including in conversations I witnessed. I also became aware from conversations
that Jody Drake and an individual named Nasser were aware that the machines were being
burned in.

13. Every machine sent to consumers was burned in before shipment. Mark
instructed me to always have as many machines burning in as possible. When production of the
mining machines was slow, we would keep machines mining all day rather than shipping them.
Machines were burned in for far longer than was necessary to test them. Generally, a machine
only needs to tested for ten to 30 minutes to be tested properly.

14. At times, we would have machines burning in for two days before shipment, in
order to generate additional bitcoins. If we reached the end of the day without additional
machines ready to replace them, we held back shipment to keep them burning in all night, and
then would ship them the next day. When production was high, the machines were burned in for

shorter periods of time.

3

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 4 of 7

15. | From my work in the shipping room and the burn in rooms, I observed the pace at
which the mining machines were produced and sent. When I arrived in June 2013, Butterfly
Labs had produced some Jalapeno 65 nm machines, but did not have working parts to mass
produce the higher powered 65 nm machines. [| estimate that Butterfly Labs would ship out 30-
50 units a day when I started, about 100 a day by the time I started constructing the burn-in
rooms in September, and about 900 by the time I completed the burn in rooms about three weeks
later. At the beginning of this time period, Butterfly Labs was shipping Jalapenos, and it started
shipping “singles” in July, and mini-rigs in September. The company shipped one rig in a week,
which was celebrated as a big deal at the company and shipped three mini-rigs by the time I left
the company. The company also sent nine machines off to Netsolus, a data center, which
operated the machines on behalf of Butterfly Labs customers.

16. ASIC 65 nm mini-rigs started shipping in the fall of 2013. The mini-rigs had
been advertised as a 1.5 TH/s machine with 1,500 watt power consumption. However, Butterfly
Labs was unable to produce a machine with those specifications. Instead of sending single 1.5
TH/s machines, Butterfly Labs shipped customers who had pre-ordered that machine three
separate 500 GH/s machines with 2,400 watt power consumption. Also, based on what I
observed from testing, the 500 GH/s machines hashed at a rate of 460 to 490 GH/s. Because the
ASIC mini-rigs had less computing power when broken up into components and consumed more
power, consumers were receiving a product that was less powerful and would generate fewer
bitcoins than they expected.

ie The three supposed 500 GH/s machines were shipped separately and at different

times. Butterfly Labs would ship the first part to a large number (and possibly all) waiting

4

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 5 of 7

customers, and then send each of the customers the second component, and then the third after
all customers had received the second component.

18. The ASIC 65 nm mini-rigs came with a Nexus 7 computer tablet that was pre-
loaded with Butterfly Labs’ EZ Miner software. Customers had to set up the mining software in
order to operate their machines. However, the default setting was to mine on the EMC pool. As
a result, the bitcoins mined by customers who incorrectly set up the machine and software went
to EMC rather than the customer. Butterfly Labs personnel, including Mark were aware that this
was happening.

19. [ purchased an FPGA mini-rig in May 2012, and upgraded it to a 65 nm ASIC
mini-rig in July 2012 by trading it in when the mini-rig started shipping and paying an additional
$15,000. I was not satisfied with the 65 nm ASIC mini-rig. Instead of coming as one machine,
it came as three separate 400 to 500 GH/s machines in separate boxes with separate power
supplies. As a result, it consumed six to seven times more power than advertised. I was one of
the first customers to get the mini-rig. The first machine arrived in July 2013, and then I had to
wait for all of the other customers in the queue to get their first machine before I got the second
machine. I did not receive the final machine until sometime in September or October 2013. Due
to the delay, it also mined far fewer bitcoins than I expected. Even being one of the first
customers to receive the machine, I barely recouped the amount I paid.

20. The company began taking pre-orders for the Monarch 28 nm machine in August
2013, but had not fulfilled all of the pre-orders for the 65 nm machines. Someone on the
management team explained that one reason that the company was taking pre-orders for the 28

nm machine was so that it could upgrade customers awaiting the 65 nm machine to the Monarch,

5

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 6 of 7

which would mean that the company would not have to produce or ship a 65 nm machine to that
customer.

21. I was terminated from the company in November 2013 because my laptop, which
contained bitcoins mined from burning in customer machines was stolen. When I left the
company, Butterfly Labs had not yet built a prototype of the machine. As far as I could tell, the

company had only created a mock-up of what the machine would look like.

Pursuant to 28 U.S.C. § 1746, 1 declare under penalty of perjury that the foregoing is true and

correct.

Case 4:14-cv-00815-BCW Document 42-13 Filed 09/27/14 Page 7 of 7